Taxonomy of Constitutional Replay-Authority A Taxonomy for Computational Legitimacy and Replay Verification

Abstract

This taxonomy establishes Constitutional Replay-Authority as a distinct new infrastructure category dedicated to computational legitimacy and replay verification. It addresses the Computational Legitimacy Gap — the structural deficiency in existing trust infrastructure whereby computational systems can record, observe, explain, govern, or preserve decisions, yet cannot reliably determine whether those decisions can later be authoritatively replay-verified under the precise conditions that existed at the moment in time they were made.

At its core, Constitutional Replay-Authority is organized around a single governing objective: “This system made THIS decision, under THIS policy, using THESE inputs — and whether that decision can be authoritatively replay-verified.”

In this context, constitutional refers to governing rules and boundaries whose satisfaction is required before replay-authoritative legitimacy may be issued. The category introduces constitutional principles, authority boundaries, and governance mechanisms specifically engineered for long-term, deterministic verifiability of computational decisions. It treats legitimacy not as an assumed property or post-hoc interpretation, but as a governed, replay-verifiable computational state.

By delineating the category’s objective, constitutional invariants, structured classifications, and clear distinctions from prior art, this taxonomy provides a foundational conceptual framework for researchers, regulators, enterprises, and technologists confronting the growing challenge of verifiable computational accountability in high-stakes AI and autonomous systems.